Direct demonstration of HIV-1-mediated high CD4 lymphocyte turnover in humans has been difficult to document, given the lack of safe agents with which to label cells in vivo (e.g., BrdU and tritiated thymidine are considered likely mutagens). The proposed study will use a cell-labeling technique - deuterated glucose D (D-glucose) - that has recently been shown to be safe in humans. This technique will be used to determine and compare the rate of lymphocyte turnover in three groups of subjects: (A) healthy HIV-1-uninfected individual (n=4); (B) chronically HIV-1-infected (>90 days), antiretroviral-naove individuals with high (>20,000, n=4) and low (<20,000, n=4) plasma HIV-1 RNA levels; and (C) individuals from Group B who after completing the first phase of this study will be required to begin combination antiretroviral therapy (n=8).